In the early days of this blog, I talked about improving the semweb's "whirlpool rap", stressing that we need a clearer way to talk about it to non-technical types. Since then, I've learned there's an even greater challenge: convincing developers that RDF and the Semantic Web are genuinely worth the training time required before they pigeonhole it and write it off as hype.

Choosing tech is critical for software developers and managers. In the last twenty-four years of wading through languages, API's, platforms, and design approaches, I've learned to bet on established winners because there's no more costly mistake than choosing tech that later (or sooner) dies on the vine.

Picking the wrong tech can mean wasting months of personal time learning something that won't be used later, either because it's not a marketable skill or because a better approach is found. For companies, picking the wrong tech can really suck, and it keeps on sucking, since new tech means a lack of qualified developers, a need to have them "learn on your dime", and long-term maintenance concerns ... "Will we be able to find qualified developers in the future if the current team disappears?"

To me, it's more important that I can find and replace developers and not have to pay for training, ramp-up time, and course corrections. I just don't like gambling on new tech.

So how do we managers choose tech intelligently? What crystal ball can we use? One approach is to stick with clear winners. When I rummage through the software section at bookstores, my goal is usually to take the temperature of what's selling, not to find some nifty new thing to learn. If I see five books on Struts, I know I'll have an easier time finding Struts developers. New tech appears on my horizon as the one solo book on the shelf ... AspectJ, for example. In a month or more, I may see three more AspectJ books. I watch these trends, and they're usually pretty accurate.

That said, I've yet to see a single RDF book at a local bookstore, and we're betting our self-funded year long project on RDF and the Semantic web.

So why are we using RDF? Would we use it again? These are the questions!